Thanks- had no idea what driver other than some sort of tweeter as the device was not listed- thanks.
I said "sub-harmonic" above- oops, harmonics and sub-harmonics of various relationships can stimulate the coupled resonator depending on a lot of things. Probably should has said any 1/2 wave multiple or whole number fraction can transfer energy to a coupled resonator or something along those lines. Helmholtz really said it far better than I ever will.
In reality ANY finite-time signal will excite a resonance. Let's say that I excite the system at some frequency OTHER than what you claim will excite it. But, since its finite in time, I have to actually turn that signal on and off, which entails a "transient" - the turn-on and turn-off can be thought of as convolutions of the "said" frequency and a pair of step responses. Now the "said" frequency will not excite the resonance, granted, but the steps, which are broad band, will.
Yes that is true but the amount of energy transferred into a coupled radiator is low with a single transient, very low unless the transient is very large. So the result of the coupled energy will likely not be audible which was the point and discussion of the earlier post. Helmholtz does say all this best.
I have read about the new version (some cool new options), I used REW for subwoofer meaurements. I use ARTA or HOLM over REW (but I might try the latest version again)...both are free too.
You really can not look at an impulse and know if a driver is good, you might be able to compare two impulses and say one is better but I do not really know. You need to take the impulse then create all the other plots from it, then you have to listen to the driver.
Even after all that the driver is only part of a complete system and even if the driver is great on its own, it may not work well in a complete system.
Im definitely not the expert here I just think you are worrying about he impulse on its own too much.
I really didn't worry about it at all until Dr. Geddes said he looks for a compact impulse response. I can't say I'd recognize a compact impulse response vs. a non compact impulse response. So as of now, looking at the impulse response is still useless to me. The graphs generated from it are not.
Thanks Doug, much appreciated.
Dan
Ah cool...I get it, I think a compact impulse would simply mean less decay. The impulse would smooth out much earlier on the time axis.
Lingering impulses do not sound great that is for sure. Of course you can not look at one impulse alone and figure it out, you have to compare impulses of like drivers and decide which has the best impulse.
I think REW calculates frequency response by averaging the whole measurement, not by calculating an impulse and letting the user set start and end gates and then calculating the FR from that. So it's measuring something slightly different than ARTA / HOLM.
REW calculates the impulse response from a log swept sine measurement (or a series of synchronously averaged sweeps if so configured) and derives all other plots from the impulse response.
What gates does it apply to the calculated impulse response? I think it uses the 'full duration' (in layman's terms, not sure how to describe this right), so it will pick up all the room effects. I haven't found a way to gate out the reflections. Maybe the new / beta version adds that capability.
I know it does something with an impulse response, because you see the GUI say something about calculating an impulse response. But I don't think it does exactly what HOLM and ARTA do (or can do). I could be wrong.
I know it does something with an impulse response, because you see the GUI say something about calculating an impulse response. But I don't think it does exactly what HOLM and ARTA do (or can do). I could be wrong.
To adjust the window durations, types or reference time click the IR Windows button, which is at the right hand end of the toolbar on V4 or on the left of the middle group of buttons on V5. The windowed part of the impulse response is shown in a different colour on the impulse response plot, and the windows themselves are also shown on the impulse response plot. Example below is from V4. There is more info in the help files, either from within REW or online here.
An externally hosted image should be here but it was not working when we last tested it.
I'm only half way thru the thread, but this little snippet may be of interest to some.
(I found the bit on dynamic compression to be interesting)
YouTube - tonmeister86's Channel
this one is definitely a puff piece, but we get a reasonable look at the HK revolving table we have all heard about
http://www.youtube.com/user/tonmeister86arrgghh, you have to click on 'infinity training on how to listen to speakers'...sorry bout that
(I found the bit on dynamic compression to be interesting)
YouTube - tonmeister86's Channel
this one is definitely a puff piece, but we get a reasonable look at the HK revolving table we have all heard about
http://www.youtube.com/user/tonmeister86arrgghh, you have to click on 'infinity training on how to listen to speakers'...sorry bout that
Last edited:
On cone resonance, has anyone looked at a woofer's CSD waterfall plot and a graph of its directivity and seen a correlation? Does anyone suspect there might be a correlation?
Thanks,
Dan
Thanks,
Dan
I would be surprised if there were a "correlation" in the sense that I think about it. That both the directivity and the CSD will go all to he11 at the same place is more of a coincidence than a correlation. The drivers WILL get too narrow to use AND it will breakup in undamped resonances in the same frequency range. These two things are related, I suppose, but the bottom line is that once ANY driver starts to bgreakup, it is no longer usable. In fact it is generally not usable to well below that frequency because of "bleed-through" of the resonant response past the crossover ( I've often wanted to try an acoustic LP to help with this problem - can you say "foam"?)
Interesting. I'd love to see some actual data on it if anyone has such a thing. Still I'm going to try and do some experimenting this week if I have the time. It is of huge importance to designing a CD speaker. It stinks that only Augerpro has both these measurements posted, but not near the detail I'm looking for in the polar response. Where I run into issues is generally at the 45 degree point and beyond.
Anyone know why some cone resonances change in frequency as you move further off axis?
Anyone know why the amplitude of the resonances seems to increase well off axis?
Dr. Geddes, are you suggesting foam for woofers? That's an interesting idea, but I'd think it would have to be pretty thick. It might also help reduce the more audible higher HDs as well. Not that they are too much of a problem. Maybe they could be with light metal cone driver?
Thanks,
Dan
Anyone know why some cone resonances change in frequency as you move further off axis?
Anyone know why the amplitude of the resonances seems to increase well off axis?
Dr. Geddes, are you suggesting foam for woofers? That's an interesting idea, but I'd think it would have to be pretty thick. It might also help reduce the more audible higher HDs as well. Not that they are too much of a problem. Maybe they could be with light metal cone driver?
Thanks,
Dan
Geddes is hinting about foam for the horns since he has a patent on doing just that with his designs.
Woofers move in and out so its not really a good idea to put foam in front of them 😉
Could you explain this better, What details are missing from Augerpro's measurements.
All he has are 0, 20, and 40 degree measurements. I need more detail for it to be useful in a final design. Even 15 degree steps would be better, but out to 90 degrees is what I'm looking for.
You would have to give the woofer some room to move, so foam would have to be suspended at least 5-6 mm away from the cone. I can't really see where placing the foam in front of the tweeter would help the woofer's break up. Hopefully that's not what he was saying.
Dan
You would have to give the woofer some room to move, so foam would have to be suspended at least 5-6 mm away from the cone. I can't really see where placing the foam in front of the tweeter would help the woofer's break up. Hopefully that's not what he was saying.
Dan
90 degrees does not matter with those woofers? You do not want to go higher on a woofer then -6dB at 45 degrees anyways so Im very curious on what you are thinking in terms of 90degrees.
Which woofer again are you thinking about?
I guess he would like to try a low pass "foam" solution. You are right.
Im building new baffles now that are 2 1/4" thick and I will mount the woofer behind. It would be easy to place some foam over the front since the woofer is recessed back. I wonder what density would matter.
Last edited:
No, I AM suggesting using foam for the woofer and NO I don't have a patent on that. I guess I'm not the blood sucking scoundral that you take me for!
I do have some highly detailed data for the woofers that I use, but not anything else.
- Status
- Not open for further replies.